Busca Resgatados
How a website developed in 24 hours helped thousands of people to find their loved ones during the most devastating flood ever seen in Southern Brazil.
- Role
- Product designer, from the idea to the interface
- Timeline
- May 2024 · first version in 24 hours
- Team
- Gabi Branco, Amanda Medeiros, Igor Carrasco and a group of volunteers
- Outcome
- 27,000 unique usersCovered by CNN Brasil (opens in new tab), O Globo (opens in new tab) and Exame (opens in new tab)
- product design
- social impact
The flood
Rio Grande do Sul, my home state, faced unprecedented floods in May this year. 581,638 people were displaced from their homes, and around 76,000 had to be rescued amidst the floods.
- people displaced
- 581,638
- people rescued
- 76,000
- to build the website
- 24h
- unique users
- 27,000
People often had to leave their houses without means of communication with their families — for several reasons, such as poor connection in the area, power outages, or simply not having a smartphone to let their loved ones know they were ok.
Volunteers working at shelters created several spreadsheets listing names of rescued and sheltered people, and started sharing those links on social media so these people could be found.
The idea
The lack of a system led to a lot of disorganized information. While trying to find a person in one of those spreadsheets, I felt desperate. I had to navigate between different spreadsheets, different tabs, unsure of how the person’s name had been written — did they use a full name? Maybe there was a typo? How was I supposed to know in which shelter the person went to or which group had rescued them?
So I understood that my Product Design skills could be useful to minimize the issues I had faced.
I reached out to Amanda Medeiros, a front-end engineer, my sister, Gabi Branco, who is an engineering manager and Igor Carrasco, a back-end engineer, and told them my idea: to build a search tool as simple as Google that would aggregate the data so people could find their loved ones.
Product requirements
We should be able to use the existing spreadsheets as our database
It would not be realistic to ask volunteers in shelters to change their workflow to feed our database. I had already done a little bit of web scraping with Python, so I knew it would be a possibility for us to consume their data.
The search mechanism should be able to handle typos
People were under a lot of stress and many names contained typos, so the tool had to accommodate that. It also should be able to give people similar results: for example, if a person is called “Maria Alice dos Santos”, we should be able to inform if we found a “Maria dos Santos”, since we couldn’t predict how those names were informed in shelters.
We had a lot of responsibility when giving a feedback
We couldn’t just say “no results found”. We were dealing with people in a very critical situation. The tool should be transparent about the spreadsheets we were consulting and about our search mechanism — encouraging people to come back later or to keep on looking for their loved ones in another database if, for instance, a particular spreadsheet hadn’t yet been integrated into our website.
Once we understood the requirements, I jumped into Figma to start prototyping. I knew that this was not the time to worry about creating the most appealing UI of my life, so I opted for Tailwind as I knew it had a reliable library that would make UI decisions easier and also would be faster to implement from a front-end perspective.
Search mechanism
The main goal was to provide a really simple interface to minimize cognitive load in a way that even first-time users could easily understand what to do. Inspired by Google Search, a tool that I assumed people would most likely know, I decided to wrap the interface around a simple search box.

We had to make sure users were aware of all the data sources our tool searched through. To solve this, we included a list of the spreadsheets we were currently integrating with, as well as those we were currently working on adding, allowing users to search these lists directly if needed.

Displaying the results
Even when a specific or similar name was found, we still had to give people resources to check if that was the person they were looking for. The solution for that was to keep every name clickable, taking the user to our original source where they could find more information.

And in case the names found weren’t a match or we couldn’t find any results, we informed that our database was being constantly updated — and also listed all of the sources we were currently consulting and had an open channel to receive new spreadsheets.
Impact
The website had 27,000 unique users and was cited on major Brazilian news platforms, such as CNN Brasil (opens in new tab), O Globo (opens in new tab) and Exame (opens in new tab). But the most important results didn’t come from any measurable key metrics, the number of requests we had on our server or how many people were simultaneously online on the website —
we knew that if at least one person could find a loved one in our platform, all our work was worth
it.
During those weeks, we received feedback that kept us going. We heard stories of families who found each other through our platform, and of social workers helping elderly people in shelters reconnect with their families — it was a small source of hope in the middle of the chaos.
Takeaway
As I finish writing this article almost four months after the tragedy, I realize there are many things we could have done differently and almost wonder if I should actually publish this article. But the most important personal takeaway for me as a designer was I had to take off my 8-point grid cop hat, blend all the frameworks I’d used throughout my career, and quickly put together a solution.
Thank you
Busca Resgatados wouldn’t have been possible without the help of these incredible people:
Gabi Branco, Bibiana Branco, Igor Carrasco, Amanda Medeiros, Douglas Eloy, Adriano Viana, Pedro Vanzella, Lucas Trajano, Henrique Barbosa, Andressa Quines, Maicon Machado, Victor Guimarães, Fernando Martinkovics, Luísa Giusti, Helena Martinelli Serra, Elisa Albuquerque, Carol Goyer, Helena Poetini, Salva RS.
Thank you so much :)
First published on Medium (opens in new tab), September 10, 2024.